Left ventricular remodeling after myocardial infarction is a major cause of subsequent heart failure and death. The capillary network cannot keep pace with the greater demands of the hypertrophied but viable myocardium, resulting in myocardial death and fibrous replacement. The first series of experiments of the present invention, described below, show that human adult bone marrow contains endothelial cell precursors with phenotypic and functional characteristics of embryonic hemangioblasts, and that these can be mobilized, expanded, and used to induce infarct bed vasculogenesis after experimental myocardial infarction. The neo-angiogenesis results in significant and sustained increase in viable myocardial tissue, reduction in collagen deposition, and improved myocardial function. The use of cytokine-mobilized autologous human bone marrow-derived angioblasts for revascularization of myocardial infarct tissue, alone or in conjunction with currently used therapies, offers the potential to significantly reduce morbidity and mortality associated with left ventricular remodeling post-myocardial infarction.
Although prompt reperfusion within a narrow time window has significantly reduced early mortality from acute myocardial infarction, post-infarction heart failure is increasing and reaching epidemic proportions (1). Left ventricular remodeling after myocardial infarction, characterized by expansion of the initial infarct area, progressive thinning of the wall surrounding the infarct, and dilation of the left ventricular lumen, has been identified as a major prognostic factor for subsequent heart failure (2, 3). This process is accompanied by transcription of genes normally expressed only in the fetal state, rapid and progressive increase in collagen secretion by cardiac fibroblasts, deposition of fibrous tissue in the ventricular wall, increased wall stiffness, and both diastolic and systolic dysfunction (4, 5). Hypoxia directly stimulates collagen secretion by cardiac fibroblasts, while inhibiting DNA synthesis and cellular proliferation (6). In animal models, late reperfusion following experimental myocardial infarction at a point beyond myocardial salvage significantly benefits remodeling (7). Moreover, the presence of a patent infarct related artery is consistently associated with survival benefits in the post-infarction period in humans (8). This appears to be due to adequate reperfusion of the infarct vascular bed which modifies the ventricular remodeling process and prevents abnormal changes in wall motion (9).
Successful reperfusion of non-cardiac tissues rendered ischemic in experimental animal models has recently been demonstrated by use of either circulating or bone marrow-derived cellular elements (10-13). Although the precise nature of these cells was not defined in these studies, the presence of precursor cells in both adult human circulation and bone marrow which have the capability to differentiate into functional endothelial cells, a process termed vasculogenesis (14-16), has been shown. In the pre-natal period, precursor cells derived from the ventral endothelium of the aorta in human and lower species have been shown to give rise to cellular elements involved in both the processes of vasculogenesis and hematopoiesis (17, 18). These cells have been termed embryonic hemangioblasts, are characterized by expression of CD34, CD117 (stem cell factor receptor), Flk-1 (vascular endothelial cell growth factor receptor-2, VEGFR-2), and Tie-2 (angiopoietin receptor), and have been shown to have high proliferative potential with blast colony formation in response to VEGF (19-22). The subsequent proliferation and differentiation of embryonic hemangioblasts to adult-type pluripotent stem cells appears to be related to co-expression of the GATA-2 transcription factor, since GATA-2 knockout embryonic stem cells have a complete block in definitive hematopoiesis and seeding of the fetal liver and bone marrow (23). Moreover, the earliest precursor of both hematopoietic and endothelial cell lineage to have diverged from embryonic ventral endothelium has been shown to express VEGF receptors as well as GATA-2 and alpha4-integrins (24). The first series of experiments of the present invention shows that GATA-2 positive stem cell precursors are also present in adult human bone marrow, demonstrate properties of hemangioblasts, and can be used to induce vasculogenesis, thus preventing remodeling and heart failure in experimental myocardial infarction.
Growth of new vessels from pre-existing mature endothelium has been termed angiogenesis, and can be regulated by many factors including certain CXC chemokines (47-50). In contrast, vasculogenesis is mediated by bone marrow-derived endothelial precursors (51-53) with phenotypic characteristics of embryonic angioblasts and growth/differentiation properties regulated by receptor tyrosine kinases such as vascular endothelial growth factor (VEGF) (54-57). Therapeutic vasculogenesis (58-61) has the potential to improve perfusion of ischemic tissues, however the receptor/ligand interactions involved in selective trafficking of endothelial precursors to sites of tissue ischemia are not known. The second series of experiments of the present invention, described below, show that vasculogenesis can develop in infarcted myocardium as a result of interactions between CXC receptors on human bone marrow-derived angioblasts and ELR-positive CXC chemokines induced by ischemia, including IL-8 and Gro-alpha. Moreover, redirected trafficking of angioblasts from the bone marrow to ischemic myocardium can be achieved by blocking CXCR4/SDF-1 interactions, resulting in increased vasculogenesis, decreased myocardial death and fibrous replacement, and improved cardiac function. The results of the experiments indicate that CXC chemokines, including IL-8, Gro-alpha, and stromal-derived factor-1 (SDF-1), play a central role in regulating vasculogenesis in the adult human, and suggest that manipulating interactions between CXC chemokines and their receptors on bone marrow-derived angioblasts can lead to optimal therapeutic vasculogenesis and salvage of ischemic tissues. The third series of experiments, described below, show that CC chemokines also play a role in mediating angioblast chemotaxis to ischemic myocardium.
The angiogenic response during wound repair or inflammation is thought to result from changes in adhesive interactions between endothelial cells in pre-existing vasculature and extracellular matrix which are regulated by locally-produced factors and which lead to endothelial cell migration, proliferation, reorganization and microvessel formation (70). The human CXC chemokine family consists of small (<10 kD) heparin-binding polypeptides that bind to and have potent chemotactic activity for endothelial cells. Three amino acid residues at the N-terminus (Glu-Leu-Arg, the ELR motif) determine binding of CXC chemokines such as IL-8 and Gro-alpha to CXC receptors 1 and 2 on endothelial cells (49,71), thus promoting endothelial chemotaxis and angiogenesis (47-48). In contrast, CXC chemokines lacking the ELR motif bind to different CXC receptors and inhibit growth-factor mediated angiogenesis (49-72). Although SDF-1, an ELR-negative CXC chemokine, is a potent inducer of endothelial chemotaxis through interactions with CXCR4 (73), its angiogenic effects appear to be limited to the developing gastrointestinal tract vascular system (50).
Vasculogenesis first occurs during the pre-natal period, with haemangioblasts derived from the human ventral aorta giving rise to both endothelial and haematopoietic cellular elements (74,75). Similar endothelial progenitor cells have recently been identified in adult human bone marrow (52-53), and shown to have the potential to induce vasculogenesis in ischemic tissues (59-62). However, the signals from ischemic sites required for chemoattraction of such bone marrow-derived precursors, and the receptors used by these cells for selective trafficking to these sites, are unknown. Following myocardial infarction a process of neoangiogenesis occurs (62,63), but is insufficient to sustain viable tissue undergoing compensatory hypertrophy, leading to further cell death, expansion of the initial infarct area, and collagen replacement (64-66). This process, termed remodeling, results in progressive heart failure (67-69). In the experiments described below, a nude rat model of myocardial infarction was used to investigate whether CXC chemokines containing the ELR motif regulate migration of human bone marrow-derived angioblasts to sites of tissue ischemia. Moreover, since selective bone marrow homing and engraftment of haematopoietic progenitors depends on CXCR4 binding to SDF-1 expressed constitutively in the bone marrow (76-78), whether interruption of CXCR4/SDF-1 interactions could redirect trafficking of human bone marrow-derived angioblasts to sites of tissue ischemia, thereby augmenting therapeutic vasculogenesis, was examined. The results of the experiments indicate that CXC chemokines, including IL-8, Gro-alpha, and SDF-1, play a central role in regulating human adult bone marrow-dependent vasculogenesis. Further, the fourth series of experiments described below show that stem cells can induce angiogenesis in peri-infarct tissue.